The application infrastructure market reflects the convergence and overlap of many of the solutions available to support application development, deployment and execution (see "'Application Infrastructure' Reflects New Dynamics in the Software Market"). The part of the market focused on integrating back-end application systems is seeing new challengers and leaders emerge, indicating that the market is incrementally maturing; however, even the leaders have more work to do to solidify recent growth, complete modernization projects and secure the long-term retention of their market shares (see Figure 1).

Application integration technology burst into the middleware market in the mid-1990s. At that time, integration products were offered exclusively by specialists. The term "specialist" is used to characterize vendors whose products and services are focused on the problem of application integration. Initially termed "message brokers," the products from these vendors focused on providing a graphical approach to specifying the business logic required to transform and intelligently route data among applications. Gradually, the features provided in conjunction with the broker expanded to the point of becoming a suite. The products offered by enterprise service bus (ESB) vendors are evolving in the same manner — adding services that will result in a suite, although ESBs are still a subset of integration suites.

Characteristics of specialist vendors include:

• A broad set of integration services.
• They typically outinnovate megavendors.
• Components of suites fit well together and have a single, comprehensive architecture.
• Products are designed for a heterogeneous environment, encompassing disparate application servers. However, nonproduct assets, such as solutions (for example, a pre-configured order-to-cash process), and knowledge-based assets (for example, patterns, templates and best practices) are not as extensive as those offered by megavendor competitors.

During the period of 1999 to 2001, IBM and Microsoft entered the application integration market, where they have since been joined by Oracle, SAP and Fujitsu. These vendors are often referred to as "megavendors." The term "megavendor" is used to characterize large vendors that provide products and services whose applicability extends beyond that of application infrastructure. Early products offered by these vendors could be characterized as "testing the waters." The solutions were more limited in scope and often contained immature or unintegrated components. Moreover, buying into megavendor products often resulted in vendor lock-in, because when an organization commits to the megavendor "stack," there are technical reasons (for example, proprietary extensions) and business reasons (for example, strategic relationships) that perpetuate follow-on sales. Therefore, products from these vendors are not well-suited to a best-of-breed approach to deploying an application infrastructure.

Megavendors' offerings are reaching functional parity with specialist offerings. Therefore, although specialists are attempting to outmaneuver megavendors by providing innovation, the megavendors are countering by expanding the integration sale to include a broad set of assets that are complementary to integration (such as solutions, services, patterns and templates) and add significant value to organizations deploying service-oriented architectures (SOAs).

Interest in SOAs has directed attention away from back-end application integration. However, integration remains a critical discipline, which increases the value of your IT portfolio and better positions that portfolio to deliver business value (see "Integration Is Out of the Limelight but Growing in Importance" to understand why integration is growing in importance).

The predominant types of software employed for the back-end application integration usage scenario are integration suites and ESBs. Competition among vendors supplying these types of products is tight. Megavendors and specialists vie for suite customers whose interest is shifting from integration suites to ESB technology.

In part, aggressive demand for ESBs is driven by companies moving to SOAs. ESBs were designed to support the deployment of SOA applications, whereas integration suites have evolved to support SOA. This does not imply that integration suites are less capable of supporting SOA deployments. Rather, it indicates that, because of costs and reduced complexity, ESBs are suited to individual projects and incremental SOA adoption.

More germane to the back-end application integration market is that ESBs provide low-cost, easy-to-use alternatives to integration suites in projects that require interfaces with simple and medium complexity. The base functionality required for those projects includes messaging, data transformation, intelligent routing, orchestration and adapters. ESBs and integration suites have these features in common; therefore, ESBs are viewed by many organizations as "good enough" for integration products.

Application integration is giving applications that were designed independently the ability to interoperate. In 1999, Roy Schulte and Jim Klein published a paper identifying three integration styles: data consistency, multistep processes and composite application integration.

The objective of data consistency integration is to make data across all applications consistent. For example, if a customer changes a billing address in a CRM application, then that event is pushed out to other applications (for example, accounting, billing and ERP) so that those applications can update their databases with the most current data.

Multistep process integration entails orchestrating the execution of the activities of a business process, regardless of whether those activities are performed by software (applications or services), humans or even intelligent devices in a manner that automates the flow of execution through instances of that business process. Comprehensive support for multistep process integration entails supporting multiple styles of business processes, including system to system, human to human, collaborative, document-centric and administrative.

Composite applications, the third style of integration, are growing rapidly as organizations seek to leverage established assets (including the services created using SOA) and minimize the amount of new code that must be developed and maintained (see "Magic Quadrant for Application Infrastructure for Composite-Application Projects, 2Q07" for composite application integration usage scenarios).

This Magic Quadrant emphasizes product capabilities most relevant to projects that have, as their primary objective, integration of the back-end data and business logic of applications. These projects deliver increased cohesion and unified access to resources of purchased packages; enterprise legacy products; and applications that are new and custom-designed from other relevant enterprises or provided by service providers.

We evaluate 24 vendors in this Magic Quadrant. Vendors not considered may also be suitable for projects in particular geographies or vertical markets, or with specific requirements. Note that while this Magic Quadrant was in production, Software AG acquired webMethods. The timing of the acquisition did not allow us to analyze the resultant entity for this Magic Quadrant; therefore, the two vendors are rated separately, although the announced acquisition influenced the rating of both vendors. Software AG has the potential of becoming a vendor with one of the richest and more technically advanced offerings for back-end integration. However, success of the acquisition is not a given because the company will face significant cultural, business, organizational and product-line rationalization challenges (see "Software AG's Proposed Acquisition of webMethods Marks Strategic Shift Toward SOA").

To succeed in the modern business computing environment, organizations must have the flexibility to experiment and innovate while preserving the overall integrity and quality of service for their core systems. To achieve this, most enterprises are engaged in two kinds of projects:

• Systematic — Designed to advance the core enterprise computing capability.
• Opportunistic — Designed to experiment with new opportunities (see "Systematic vs. Opportunistic: Useful Heterogeneity").

The product and vendor selection decisions considered in this Magic Quadrant target the systematic projects and their holistic view of the enterprise's application infrastructure.

Inclusion in this Magic Quadrant is based on an assessment of the functional capabilities of products that were generally available in the market at the beginning of 2007 (see "'Application Infrastructure' Reflects New Dynamics in the Software Market"). We have assessed the relevance of these capabilities to the data consistency and multistep process integration styles. The capabilities considered most important include interoperable messaging, "classic integration" services and orchestration. The importance of these capabilities was incorporated into the set of weights used when computing the scores for the back-end application integration usage scenario.

We assessed each supplier's products with respect to these capabilities, and used that assessment in combination with the weights to create an overall score for each supplier. That score was used to determine which suppliers would be included in the Magic Quadrant, and it was also used as the score for the product/service criterion in the Magic Quadrant. The 24 suppliers included are the result of our assessment of those most relevant for consideration for this usage scenario.

The ability to execute in the application infrastructure for the back-end application integration usage scenario primarily reflects maturity and completeness of the product offering, reasonably affordable costs and the presence of a sizeable customer base that reports good experience with the vendor's integration technology. Vendors' responsiveness in reacting to market trends, ability to communicate a sensible value proposition and effectiveness in supporting customers with a proper organization are also important factors, although not more so than for other markets. In other words, product innovation and quality, users' productivity, vendors' commitment to client success and low costs are heavily weighted; the prospects for product survival have a moderate weighting; and vendor viability has the least weighting of these factors (see Table 1).

Technology providers are evaluated on their ability to convincingly articulate logical statements about market direction, innovation, customer needs and competitive forces, and how well they map to Gartner's position. Ultimately, technology providers are rated on their understanding of how market forces can be exploited to create opportunities for providers (see Table 2).

Leaders are vendors with a proven and comprehensive integrated product set, as well as a sizeable installed base of reasonably satisfied clients. They have an ample installed base of products to cross-sell their integration solutions to, and have demonstrated their ability to anticipate technology and market trends over the years by extending their offers with composite, application-enabling technology.

Leaders manifest their understanding of the problem space by providing innovative features and extending their integration technology to support initiatives, such as SOA and business process management (BPM).

Challengers are vendors that have demonstrated their technology can support the implementation of numerous, large, business back-end integration projects, and have built platforms capable of effectively competing against, and often besting, those of the leaders.

However, these vendors are followers rather than leaders in providing new, innovative features, or their innovation is focused on a specific problem space. In some cases, their vision is not manifested through focused marketing messages and value propositions. Most of the challengers have the opportunity to become leaders through greater product innovation, combined with marketing and sales focus on back-end application integration.

Visionaries demonstrate innovation from a sales and marketing perspective. They have significant investments in integration technology, and their prospects for survival and growth depend on their ability to establish a strong presence in application infrastructure for back-end application integration. However, products of some of the larger visionaries have relatively small installed bases and, in some cases, their production readiness is not fully proven via a spectrum of mission-critical deployments.

Through diligent and focused execution, some of these players have the opportunity to become leaders. Other vendors are small and have limited sales, marketing and support resources, creating significant obstacles in the pursuit of their ambitions. Many of these vendors are likely to merge or be acquired by larger companies, but some offer excellent and highly innovative products that outperform large vendors' offerings.

The best products in the Visionaries Quadrant can be used together with point products from other vendors to create top-quality middleware infrastructures that have all the features offered by the "one-stop shopping" suites of larger vendors in the Leaders Quadrant.

Niche players usually offer good and, in some cases, excellent back-end integration technology. The focus of some niche players on a specific vertical market has resulted in products that are less useful in integration problems outside that domain. Other vendors lack focus on this problem space, which, for them, is a marginal business. Some vendors also have limited sales, marketing and support resources, and others are committed to only one geography or installed base.

Nevertheless, the back-end integration technology from these vendors can be an optimal choice for specific classes of users (for example, users in a particular vertical market where the vendor's integration technology is focused, or users in the same geography where the vendor is focused). Some vendors in this quadrant could emerge as visionaries through a greater commitment to innovation and focus on this market.

• Biggest integration and business-to-business (B2B) vendor in Europe.
• Well-established customer base worldwide.
• Mature technology; well tested.
• Strong system integration parent (Sopra) and senior management.

• Limited Axway brand awareness.
• Resolving the duplication in functionality coming from the recent Axway acquisitions within Synchrony, minimizing customer impact.
• Historically low, average size of deal (although this is improving with time).
• Counter the threats posed by competing with megavendors.

• Strong, traditional integration offering (WebLogic Integration), including a large set of OEM-sourced adapters.
• Combined offering of Java-platform-optimized integration technology (WebLogic Integration) and SOA-optimized ESB integration technology (AquaLogic Service Bus).
• Recent, successful acquisitions of leading visionary vendors in adjacent areas, including a visionary BPM vendor, Fuego.
• Leading presence and reputation in high-end enterprise computing, as well as vertical market (including telecommunications, financial services and government) and international (including a leading position in the People's Republic of China) strength and name recognition.

• Most integration technology depends on WebLogic Server and favors Java.
• Relatively small ESB (AquaLogic Service Bus) and data integration (AquaLogic Data Services Platform) production installed base.
• Packaging WebLogic Integration and AquaLogic Service Bus reflects user preferences in functionality, but increases costs to users and prospects.
• No dedicated tools to support batch processing.

• Fast-growing business (more than 60% in 2006) and installed base (300 customers).
• Rich Cape Clear 7 ESB platform — including ESB-core, transformation, routing, microflow, orchestration, management, security, business activity monitoring (BAM) and development tools — that appeals to users looking for a pragmatic, fast-to-deploy, back-end integration platform.
• Highly scalable and available BPEL-based microflow and orchestration technology.
• Focuses on providing technology and architecture specifically designed to enable "integration as a service."

• Limited brand recognition vs. other back-end integration technology vendors.
• Sales activity is concentrated in North America and the U.K., with only a small presence in continental Europe and Asia/Pacific.
• Lack of service registry/repository, human workflow and B2B offerings limits Cape Clear's appeal to users looking for an all-in-one, back-end integration platform.
• Its integration-as-a-service-oriented business model is still unproven.

• Advanced, event-driven internal architecture and outstanding high-availability, high-performance engineering for middleware, including integration middleware.
• Well-integrated product that is strong in high-productivity design tools and advanced deployment infrastructure.
• Built-in ESB, BPM, XSLT and JCA-based external integration.
• Significant, private financial resources.

• Limited support for data reconciliation, B2B and other traditional forms of back-end integration.
• Relatively small and slow-growing total installed base, which is not in the back-end integration space.
• Product positioning does not emphasize back-end integration.
• Lack of significant partnerships (apart from WebEx and some system integrators in the Netherlands), which are not in the back-end integration space.

• Unique and highly productive development process and runtime paradigm.
• Proven, scalable and reliable message-oriented middleware (MOM) backbone.
• Compatible with diverse operating systems and application servers.
• Able to address a broad range of projects from simple SOA to high-end, sophisticated application integration scenarios, including those requiring orchestration.

• Low visibility in the market.
• Small vendor with limited sales and support staff for SOA and the application integration market.
• Supplies only the SOA and integration backplane; must partner to deliver BAM, BPM, and portal and workflow capabilities.
• Must stretch development resources to stay current with evolving industry standards.

• Deep understanding of back-end integration requirements, with a long history of supporting the mainframe environment.
• Features address batch-based and real-time integration requirements.
• A growing number of deployments inside and outside the U.S. that integrate mainframe legacy applications (such as COBOL applications on a mainframe) with applications in an open-systems environment.
• Comprehensive product line in SOA and BPM supported by an expanding set of consulting resources.

• Lack of a track record integrating major packaged applications.
• Lack of a presence and track record for Interstage Service Integrator and Interstage Business Process Manager.
• Limited presence in the mainframe environment outside of Japan.
• Lacks the articulation for its products and the benefits of application portfolio analysis capabilities that assist in leveraging established assets and integration opportunities for users.

• Hitachi hardware (storage) brand recognition is leveraged for the sale of Cosminexus, Hitachi's integration suite.
• Hitachi's methodologies and track record in supporting mission-critical systems attract users.
• Hitachi is expanding its service offerings, which are based on a practical methodology.
• Good track record of supplying high-quality products backed with high-quality support

• System integration and support resources for Cosminexus are limited outside Japan.
• Poor articulation of vision, product strategy, usage scenario and value proposition for back-end integration projects.
• History of past U.S.-based project failures is impeding sales.
• Hitachi is not evolving Cosminexus through innovation, and the product continues to lose ground to back-end application integration market leaders.

• Best-of-breed adapter portfolio, plus a moderately good integration suite.
• Suitable for all integration patterns, including SOA.
• Global sales and support.

• Architectural coherence.
• Reliance on third-party MOM and SOA governance.
• Adapter commoditization and channel partner conflict.

• IBM brand recognition, global reach and "mind share," and a huge installed base of products that are leveraged for WebSphere sales.
• Comprehensive product line for integration and SOA that includes BPM, complex event processing and BAM.
• IBM Global Services — the Business Computing Services division — support with a comprehensive methodology for SOA.
• Large number of strategic partnerships with independent software vendors (ISVs) that base their products on WebSphere technology.

• WebSphere Process Server was launched in September 2005, but there are only 300 customers and 20 production deployments as of April 2007.
• When customer requirements include demanding performance and connectivity, as well as transformation in a heterogeneous IT environment, IBM recommends combining Process Server with WebSphere Message Broker — an additional product (and cost).
• For implementing complex, back-end interfaces in a heterogeneous environment, IBM is likely to suggest the following products: WebSphere Modeler, WebSphere Integration Developer, WebSphere Process Server, WebSphere Message Broker and WebSphere Monitor. Some users have reported difficulty in configuring, deploying and managing this diverse set of products.
• The pricing for the large set of products listed above makes them unaffordable to projects in mainstream and conservative organizations that are starting to use integration technology.

• Leading presence and name recognition in healthcare markets as an integration technology provider; strong support of healthcare industry integration standards and protocols.
• Many successful partnerships for integration solutions with healthcare and other ISVs.
• Well-designed, well-integrated, high-performance technology suite with a track record in multiple patterns of application integration and high productivity for integration projects.
• Profitable and well-run private business without debt that delivers significant financial resources.

• Modest name recognition outside the healthcare market (although its presence in a few other vertical markets is growing).
• Proprietary programming model requires longer learning time than typical .NET frameworks or Java-based integration technologies.
• Minimal influence in industry standards and consortia outside of healthcare.
• Privately held business that is prone to surprises more often than transparent, publicly owned businesses.

• The company regained profitability and growth, owing to its singular focus on SOA and ESB technology.
• Dual and complementary open-source (Celtix Enterprise) and closed-source (Artix) ESB strategy.
• Fast-growing Artix business and customer base of large deployments (about 100).
• Rich and extensible Artix-based, back-end integration offering, including ESB, orchestration, transformation and metadata management. Compatible with open-source Celtix Enterprise.

• IONA's and Artix's market awareness is still lower than that of leading competitors.
• Back-end integration offering is not as rich as those of leading competitors due to the lack of event management, BAM, B2B and workflow.
• Still-unproven, open-source-based Celtix Enterprise technology. Relevant business models are still untested on a large scale.
• Limited, albeit growing, support from global and regional system integrators.

• Rich and easy-to-use iBolt integration platform based on a proven, metadata-driven application development and runtime container (eDeveloper) that provides transformation, routing, human workflow, modeling, composition, portal and BAM.
• Specific iBolt packaging addresses SAP BusinessOne back-end integration scenarios.
• Large and global network of partners (ISVs, value-added resellers and system integrators).
• Singular focus on small and midsize businesses (SMBs), and SAP, Oracle's JD Edwards World and IBM's System i ecosystems via indirect channels.

• Magic is going through management changes, reorganization, divestiture of peripheral businesses and strategy redefinition after being acquired by a larger organization (Emblaze).
• Static business in terms of revenue (2006 growth was 1% vs. 2005) that only recently regained profitability and growth.
• Limited brand recognition in the back-end integration space.
• Focus on SMBs limits its appeal for large enterprises, which are the biggest consumers of back-end integration technology.

• Brand recognition, global reach, "mind share" and huge installed base of products that are leveraged for BizTalk Server sales.
• BizTalk Server installed base of more than 6,000 customers — two-thirds are estimated to be BizTalk Server 2006 Enterprise Edition.
• Affordable pricing on projects for midsize enterprises.
• Ability to attract large numbers of ISVs that provide a wide variety of solutions based on BizTalk Server.

• Applicability is limited to the Windows environment, which means customers have integration challenges if they want mix-and-match products on other platforms to create a unified application infrastructure for a heterogeneous environment.
• Microsoft reacts to most Java-related standards rather than getting out ahead of them — for example, Java Business Integration (JBI) and service component architecture/service data object.
• Microsoft's tools are implemented in a manner that encourages opportunistic development. Integration — in particular, the integration required to support services for mission-critical applications — is best approached systematically.

• NEC's hardware ("supercomputer") brand recognition is leveraged to articulate messaging about mission-critical capabilities.
• Commitment to and track record of being among the first to implement open standards (such as BPEL 2.0, JBI and UDDI 3.0).
• Commitment to telecommunications is supported by broad capabilities in this vertical market. 
• CEO commitment to increase focus on application infrastructure.

• Limited installed base of ActiveGlobe integration products.
• Narrow set of products that lack a coherent architecture.
• Weak articulation of business value.
• Limited investment in system integration and support resources for NEC ActiveGlobe outside Japan.

• Oracle brand recognition, global reach and "mind share," and a huge installed base of products that are leveraged for Fusion Middleware sales.
• BPEL PM is a highly regarded and successful technology.
• Comprehensive product line for back-end integration and SOA that includes B2B, BPM technology, BAM, ESB and data integration.
• Cohesiveness of the back-end integration product suite (SOA Suite) via common metadata management.

• Oracle has lower, although growing, market awareness than its top competitors in back-end integration.
• Its growing product portfolio makes it difficult to retain architectural coherence.
• New Oracle ESB technology has a smaller installed base and fewer production deployments than established competitors.
• Managing transition of its legacy integration platform's (Oracle Interconnect) installed base to Oracle ESB.